On Sat, Jan 08, 2005 at 08:08:38AM +0000, Aaron Denney wrote:
I suppose I wouldn't be too upset at using the locale information, but defaulting to UTF-8, rather than ASCII for unset character set information.
But if we default to a UTF-8 encoding, then there could be decoding failures when attempting to read a file. You have to consider both the possibility that characters aren't expressible in a given encoding and the possibility that files aren't expressible in a given encoding. ASCII (or iso-whatever) has the advantage that at least every file is readable. But as you say, really what we need is a binary IO system first, and then the character-based IO can do whatever it likes without breaking things (since it'll only be used by programs that actually want unicode coding). -- David Roundy http://www.darcs.net